Learn firsthand about OpenStack, its challenges, opportunities and market adoption. My goal is to interview all 24 members of the OpenStack board, and I will post these talks sequentially at Dell TechCenter.
Q: Can you please introduce yourself?
Tristan Goode: I am the founder and CEO of a managed service OpenStack provider with presence in Australia and India. I am also the founder and lead organizer of the Australian OpenStack User Group and a supporter of the Indian OpenStack User Group, both personally and through my company, Aptira. I’m also very fortunate to be twice elected to the OpenStack Foundation Board of Directors as a representative of the individual membership. I’m also very proud to recently have become an OpenStack Ambassador.
Q: What are your duties as a board member?
A: I feel my primary duty is representing the users, operators and community builders in the OpenStack eco-system. There’s good representation on the board for vendors and developers of OpenStack so I’ve always felt that there needs to be a place for those who are involved in the OpenStack community who don’t write software or push a distribution.
Q: What is your view on the current Havanna release? What are the major enhancements compared to Grizzly?
A: The current release brings OpenStack a lot closer to being production ready out of the box. There are currently some areas of concern with regards to scalability and stability with some of the newer incubated projects but these are being worked upon and lot of effort is being put into making them as reliable as some of the older projects. Havana features a lot of improvements and bug fixes over the Grizzly release. There have been some long awaited enhancements to the Dashboard, like the ability change password as an example of a very simple and long needed function. Global clusters support for OpenStack Swift; API enhancements, improved support for Nova Cells and the addition of Docker are some major enhancements.
Q: What is your favourite feature in OpenStack Havana?
A: Usability improvements to the Dashboard and the better integration of Keystone and Neutron functionality to the Dashboard have to be up there.
Q: Is there anything that still needs to be improved in OpenStack or doesn’t work well?
A: Despite the hard work by the documentation team, the nature of OpenStack’s release cycle makes it difficult for them to keep pace with all the changes and features introduced in the new releases. More needs to be done to keep the documentation current and I want the Foundation to really focus on how we do this. Greater attention needs to be paid with some of the newly incubated projects and OpenStack’s core design tenets. More care needs to be taken to ensure that some of the new features being introduced do not break the design tenets. Also, we don’t want OpenStack to become a trojan horse for vendors so that means each project (Compute, Storage, Networking especially) should have a first-class and free option for systems administrators to implement.
Q: Let me go back to my previous question. Are there any projects in particular which you feel are not fully mature yet?
A: The Neutron Project has some scalability issues that I am a bit concerned about. While there are a lot of contributors to the Neutron project, it seems that the large chunk of the development is done on vendor specific drivers and interfaces. Currently a Neutron solution deployed to scale requires a 3rd party proprietary software, an open source solution needs to be bolstered to avoid this.
Q: How do you envision OpenStack two or three years from now?
A: As the default cloud operating system. I have a lot of conversations with enterprises that have no other choice than to “glue” various components of their infrastructure into a cohesive state using OpenStack – right across their entire business. OpenStack is the unique opportunity for nations and businesses around the world to develop ground-up cloud solutions without proprietary or data sovereignty constraints.
Q: What is your view on Linux as a road model for OpenStack?
A: The important job of the foundation is to define core very, very well because if we don’t, it’s going to lead to a fractured ecosystem. Taking Linux as an example, we need to control our OpenStack “kernel” with a rich environment and a whole bunch of different distributions around it. As an example, we need to control the use of the word certification and similar wording used with OpenStack. In many cases presently, some organisations offering training have so called “certifications” for largely proprietary flavours of OpenStack – and this has to be controlled so the consumer knows what they are getting.
Q: What are the biggest challenges ahead of the board?
A: Getting better User and Operator feedback to the development teams is critical. With the proliferation of new projects there are concerns being raised about the lack of cohesiveness between the projects and the perception that the developers are not heeding real world deployment feedback and are racing to add features and new projects. Essentially, we all need to work together and constantly look at the big picture view of the project so OpenStack is stable, scalable and secure, even that means decelerating features.
Q: Thank you for your time.
A: Thank you.